PROJECT SUMMARY Lung cancers are the major source of metastasis in the central nervous system (CNS). There is an important gap in our understanding of how brain metastases respond to therapies and what mechanisms sustain metastatic tumors in the CNS. Historically, the blood brain barrier has been viewed as an impediment to systemic drugs, and novel brain penetrant agents such as the mutant EGFR inhibitor osimertinib have been developed. However, despite improved clinical responses with these agents, brain metastases still progress, and it is unknown how perturbations in the brain tumor microenvironment (TME) can be leveraged for more effective treatments in patients with CNS disease. We have developed novel methods to molecularly characterize human cerebral spinal fluid (CSF) as well as distinguish tumor from stromal gene alterations of brain metastasis in vivo. Our approaches uncover genetic mutations as well as brain TME induced alterations that converge onto cooperating pathways, such as those regulated by VEGF, NOTCH, ?-catenin and PI3K. We hypothesize that these molecular alterations: 1) cooperatively drive NSCLC brain metastasis and drug resistance in the brain, 2) are clinically actionable, and 3) are more accurately detected in human CSF or brain biopsies, due to divergent genetic evolution and TME induced adaptation of brain metastasis. Our hypothesis will be studied in 3 independent yet complimentary aims. In Aim1, we propose to collect human CSF from craniotomies as well as lumbar punctures of lung cancer patients with brain metastases who will be undergoing a bronchoscopic biopsy. By comparing the mutational landscape of matched CSF, plasma and tumor tissue, we will molecularly characterize humans with asymptomatic brain metastasis. Moreover, we will use novel orthotopic patient derived xenograft models (PDXs) to determine if brain metastasis progression and drug response correlates with co-occurring mutations identified in human CSF. In Aim 2, we will test the novel hypothesis that an activated brain microvasculature enhances the survival of drug resistant tumor cells via stromal induced NOTCH signaling in vivo. We will assess if novel bi-specific agents which simultaneously inhibit VEGF and NOTCH can delay brain metastasis progression and/or improve osimertinib response in pre-clinical models. Using human biospecimens, we will correlate the expression of VEGF and NOTCH pathway components with brain metastatic relapse. In Aim 3, we will conduct a clinical trial combining a mutation specific TKI (osimertinib) with a brain vascular targeting agent (bevacizumab) in treatment nave lung cancer patients with EGFR mutant tumors and CNS disease. Finally, molecular markers (including those studied in Aims 1 and 2) of response or resistance to this combination will be identified by analyzing CSF, plasma and tumor biopsies. This proposal will help uncover the biological basis of brain metastasis relapse. Importantly, our study will generate insight as to how current and prospective therapies can be harnessed to target both tumor specific mutations and the TME, in a manner that improves clinical outcomes for lung cancer patients with CNS disease.